Abstract:Micron-range untethered, magnetic helical robots have great potential for biomedical applications due to their desirable performance with high flexibility and accuracy in unstructured and confined environments. However, at the microscale, time-varying uncertain disturbances in the environment and electromagnetic system greatly hinder helical microrobot tracking control performance. When a microrobot is replaced or even a derivative version with a slight helical body structure change is used for different tasks, the performance of the original control scheme remarkably decreases or even becomes ineffective. Here, we propose a data-driven optimal integrated controller (D2-OIC) that realizes precise tracking and transfer control among a series of helical microrobots with derived structures in different situations. The control approach has a parallel structure with nonlinear feedforward and linear feedback controllers. The nonlinear feedforward controller inversely maps the relationship between the electromagnetic field state and the helical microrobot motion state, allowing the helical microrobot to quickly approach the desired motion state. The linear feedback controller effectively adjusts the controller parameters using the virtual reference feedback tuning (VRFT) method, thus eliminating any residual motion errors arising from nonlinear control. By retraining on newly acquired and collected cumulative data with assigned weights, the nonlinear feedforward controller is updated to achieve transfer control among various helical microrobot types. In the experiment, two helical microrobot types performed arbitrary path tracking and obstacle avoidance tasks with tracking errors consistently less than 4% of the microrobot body length, demonstrating the feasibility of the proposed method.

Dynamic Obstacle Avoidance for Magnetic Helical Microrobots Based on Deep Reinforcement Learning

Autonomous 3D positional control of a magnetic microrobot using reinforcement learning

Deep Reinforcement Learning Framework-Based Flow Rate Rejection Control of Soft Magnetic Miniature Robots

Intelligent Navigation of a Magnetic Microrobot with Model-Free Deep Reinforcement Learning in a Real-World Environment

Dynamic Modeling of Magnetic Helical Microrobots

Efficient Model Learning and Adaptive Tracking Control of Magnetic Micro-Robots for Non-Contact Manipulation

Hierarchical Planning with Deep Reinforcement Learning for 3D Navigation of Microrobots in Blood Vessels

Vision-Based Automated Control of Magnetic Microrobots

Data-Driven Parallel Adaptive Control for Magnetic Helical Microrobots With Derivative Structure in Uncertain Environments

Dynamic Obstacle Avoidance Technique for Mobile Robot Navigation Using Deep Reinforcement Learning

Dynamic Simulation-Guided Design of Tumbling Magnetic Microrobots

3D Navigation Control of Untethered Magnetic Microrobot in Centimeter-Scale Workspace Based on Field-of-View Tracking Scheme

Control of Magnetic Surgical Robots With Model-Based Simulators and Reinforcement Learning

Magnetic-Controlled Microrobot: Real-Time Detection and Tracking through Deep Learning Approaches

Magnetic-Field-Inspired Navigation for Robots in Complex and Unknown Environments

Hierarchical deep reinforcement learning controlled three-dimensional navigation of microrobots in blood vessels

Micro/Nano Motor Navigation and Localization via Deep Reinforcement Learning

Robust 3-D Path Following Control Framework for Magnetic Helical Millirobots Subject to Fluid Flow and Input Saturation

Enabling microrobotic chemotaxis via reset-free hierarchical reinforcement learning

Precise Control of Magnetic Soft Microrobot in Flowing Environment

Model-Based Reinforcement Learning for Ultrasound-Driven Autonomous Microrobots